Neural Mechanisms of Perceptual Inference

Rudiger vonder Heydt

Q: What percentage of the recorded cells were in V1 ?

A: 50

Q: Are these unbiased selection of random cells?

A: Yes, as we physiologists do.

Q: How is the response defined?

A: Its the total number of spikes in that window. (800ms).

Q: What is the largest extent of the visual field you tried?

A: 8 degrees.

Q:How do you measure the intitial receptive field of a cell?

A: We use an optimal bar. We move the bar out and figure out the point where this cell does not respond any more. The minimum region outside which you don't get a response.

Comment: Other techniques can give different receptive field sizes.

Q: Figure ground and ownership?

A: Its just different words for the same thing.

Q: Didn't understand the comment about influence of IT.

A: We have the half-strength of the ownership signal at 70ms. IT, upto 100ms is silent.

Q: What about the illlusory contours? ARe they more local?

A:

Q: Does that mean that the illusory contour is withing the summation field of the neuron?

A: What we know as the summation field is quite controversial. Summation os not an explantion of the illusory contours.

Q: What about natural scenes?

A: I suspect that some of this would hold up.




Back to the main index for Inference and Prediction in Neocortical Circuits.